Safety Science 101 (2018) 197–208
Contents lists available at ScienceDirect
Safety Science journal homepage: www.elsevier.com/locate/safety
Measuring work-related psychosocial and physical risk factors using workplace observations: a validation study of the “Healthy Workplace Screening” ⁎
Anne Tomascheka, , Sarah S. Lütke Lanferb, Marlen Melzerc, Uwe Debitza,1, Gabriele Burucka, a b c
MARK
⁎
Department of Psychology, Technische Universität Dresden, 01062 Dresden, Germany Department of Psychology, Albert-Ludwigs-Universität Freiburg, Engelbergerstraße 41, 79085 Freiburg, Germany Federal Institute for Occupational Safety and Health (BAuA), 01072 Dresden, Germany
A R T I C L E I N F O
A B S T R A C T
Keywords: Observation-based assessment Physical risk factors Psychosocial risk factors Musculoskeletal complaints Mental health
Occupational health research has demonstrated that work-related risk factors affect employees’ physical and mental health, performance and ability to work. In order to design healthy workplaces, a valid and comprehensive assessment of potential work-related risk factors is needed. Currently, observation-based methods are scarce, even though they would provide a meaningful complement to self-report instruments. The present study aimed at validating an observational interview, the “Healthy Workplace Screening” (Screening Gesundes Arbeiten, SGA), which measures work-related psychosocial and physical risk factors. We collected a sample of 641 SGA profiles representing various jobs and occupational settings to test construct and criterion validity. Results regarding construct validity showed medium-sized correlations with the stressor subscales of an established self-report job analysis instrument (SQUAW). Providing support for the criterion validity, jobs with varying risk profiles in SGA dimensions significantly differed with regard to mental health and musculoskeletal complaints. In sum, the SGA can be recommended as a valid and efficient observation-based instrument to identify critical work-related risk factors. Together with its suggestions for work design, it can serve as an easy to apply tool for workplace health promotion.
1. Introduction According to the European Survey of Enterprises on New and Emerging Risks (ESENER- EU-OSHA, 2012), mental health complaints such as stress, depression or anxiety are the second most frequently reported work-related health problem after musculoskeletal diseases. Work-related psychosocial factors have been shown as major contributors to mental health problems (for longitudinal studies see e.g. Ahola et al., 2006; Harvey et al., 2017; Milner et al., 2017; Niedhammer et al., 2015; for recent systematic reviews see Rau and Buyken, 2015; Theorell et al., 2015). Furthermore, a majority of findings (for longitudinal studies see e.g. Campbell et al., 2013; Warren et al., 2015; for reviews/meta-analysis see Bernal et al., 2015; Harvey et al., 2017; Hauke et al., 2011; Lang et al., 2012; Nixon et al., 2011) also highlight the importance of psychosocial factors at work (e.g. job demands, job
⁎ 1
control, feedback, social support and leadership) for the development of musculoskeletal complaints (e.g. low back pain). Consequently, the results of ESENER 2 confirm psychosocial risk factors as the biggest challenge for businesses in Europe (EU-OSHA, 2015). To understand how those work-related risk factors - both psychosocial in addition to physical – can have an adverse impact on employees’ physical and mental health the “Cinderella” – Hypothesis offers an interesting psychophysiological explanation. 1.1. “Cinderella” - Hypothesis In the nineties, the term “Cinderella” - Hypothesis was coined by Hägg (1991) for his assumption that the development of neck/upper limb musculoskeletal disorders (MSD) is caused by prolonged physical activation of low threshold motor units. Referring to the fairy tale
Corresponding authors at: Department of Psychology, Institute of Work, Organizational and Social Psychology, TU Dresden, 01062 Dresden, Germany. Present address: Hubrich and Roitzsch Partnership, Chemnitzer Str. 119, 01187 Dresden, Germany.
http://dx.doi.org/10.1016/j.ssci.2017.09.006 Received 29 June 2017; Received in revised form 25 August 2017; Accepted 10 September 2017 0925-7535/ © 2017 Elsevier Ltd. All rights reserved.
Safety Science 101 (2018) 197–208
A. Tomaschek et al.
responsibility, leadership, communication and conflicts (Demerouti, 1999; Nachreiner, 2008; Rau et al., 2010; Theorell and Hasselhorn, 2005; Waldenström and Harenstam, 2008). Waldenström and Harenstam (2008) offer the explanation that “external assessments and self-reports had partly different theoretical roots, and were thus intended to measure partially different aspects of the same dimension” (p. 248). Self-reports assess work-related risk factors as they are perceived by the worker. Therefore, they are influenced by additional factors such as personal dispositions, mood, expectations and previous experiences. Whereas observation-based instruments rather measure risk factors as they can be observed. Thereby, some issues of the work environment are rather difficult to observe (e.g. role conflict, Frese and Zapf, 1999) or “unlikely to surface in the presence of an observer” (Rugulies, 2012). This might apply to the more socially transmitted risk factors (e.g. leadership, communication) where incumbents have deeper insights and knowledge. Rau et al. (2010) argue that the inconsistence between observation-based and self-report measures of job demands is due to the conceptualization of the “construct of job demand itself” (p. 90). Defined as a pressure to manage the job well, job demands refer to an outcome associated with an inherent subjective component not a work-related risk factor per se. As a consequence job demands are difficult to observe. However, conditions (e.g. task conflicts, work interruptions, multitasking, etc.) leading to high job demands are observable and might therefore be more appropriate for observation-based measurements. In consideration of the specific challenges inherent in both measurement types, simultaneous use of observation-based and self-report instruments seems a promising strategy to accurately identifying work-related risk factors (e.g. Grebner et al., 2005; Rau et al., 2010; Schuller et al., 2012; Semmer et al., 2003; Theorell and Hasselhorn, 2005). Some comprehensive (ISTA: Grebner et al., 2005; TBS: Hacker et al., 1995; VERA/RIHA: Leitner and Resch, 2005; REBA: Richter et al., 2009) as well as sector-specific (Stab et al., 2016) observation-based job analysis instruments have been developed and validated by German-speaking research groups (for an overview of some European instruments see Tabanelli et al., 2008). Those instruments provide an in-depth analysis of work-related risk factors at a very complex and detailed level of measurement. Consequently, observer ratings are based on an extended analysis period (up to one work day) and should be realized by trained work and organizational experts (ergonomists or psychologists). However, to our knowledge, effortless and time efficient observation-based instruments are still rare. Moreover, there are no observation-based instruments that sufficiently assess both psychosocial and physical risk factors at work. Hence, the present article aims at the introduction and validation of an observer-based screening instrument which comprises both psychosocial and physical factors: The “Healthy Workplace Screening” (Screening Gesundes Arbeiten, SGA - Buruck et al., 2015; Debitz et al., 2014). The SGA represents a user-friendly, low-threshold tool for workplace risk assessment. Relying on the extended “Cinderella” Hypothesis (Melin and Lundberg, 1997), it allows a combined measurement of both psychosocial and physical risk factors. Several pilot studies (e.g., Buruck et al., 2007; Horváth et al., 2015; Keller et al., 2010) supported the validity and practicability of the SGA, a comprehensive validation study of this instrument has not been conducted yet. Therefore, the current study pursues two research goals: First, an investigation of the construct validity of the SGA instrument (correlational analysis) and second, a test of the criterion validity of the SGA instrument by analysing the relationship between the SGA dimensions and health-related outcomes (mental health, musculoskeletal complaints, and back pain).
character Cinderella, who had to work the whole time, those muscle fibres are recruited first at the onset of muscle activation and are firing non-stop until the muscle is completely relaxed. Consequently, sustained engagement leads to exhaustion and overload of those muscle fibres independent of the absolute level of force and contributes in the long run to degenerative processes as well as pain development. However, research showed that physical activation merely explains 33% of the development of all MSD (see for instance Kuhn, 2007). Therefore, Melin and Lundberg (1997) expanded the “Cinderella” Hypothesis by adding psychosocial aspects of work tasks as potential risk factors for musculoskeletal complaints. Specifically, several experimental studies (e.g. Lundberg, 2002; Rissen et al., 2002; Sjøgaard et al., 2000) indicate that psychosocial factors contribute to the prolonged recruitment of the same low threshold motor units as physical activation. Additional support for the extended “Cinderella” - Hypothesis comes from occupational health research (for longitudinal studies see e.g. Esquirol et al., 2017; Kouvonen et al., 2016; for reviews/metaanalysis see e.g. da Costa and Vieira, 2010; Nixon et al., 2011; Taylor et al., 2014; van den Berg et al., 2009) showing that both kinds of workrelated risk factors (physical and psychosocial, respectively) are associated with self-reported musculoskeletal complaints as well as impaired mental health. Accordingly, a study using data provided by the German pension fund (Deutsche Rentenversicherung) demonstrated that the relative risk of early retirement increased by 67% as a result of a combination of physical and psychosocial risk factors compared to the exposure of physical risks only (Sigrist and Dragano, 2007). Despite this evidence suggesting the importance of monitoring physical and psychosocial factors in the work environment and the corresponding policy initiatives in the EU as well as at the national level in many countries, there is still a lack of transfer into practice (EUOSHA, 2012). This might be due to a shortage of validated and userfriendly instruments for the combined assessment of psychosocial and physical risk factors at work. 1.2. Observation-based job analysis To date, the majority of job analysis instruments assessing workrelated risk factors have been based upon self-reports rather than on workplace observations (Gebele et al., 2011; Grebner et al., 2005; Leitner and Resch, 2005). There are validated comprehensive surveys (Pejtersen et al., 2010) as well as symptom-specific questionnaires (Brom et al., 2015). However, even longitudinal designs are not able to rule out the major concern of common method bias inherent in perceptual measures (Podsakoff et al., 2003). Observation-based job analysis instruments represent one valuable alternative for measuring work-related risk factors independently of the reports of the incumbents (for other approaches see Daniels, 2006; Rugulies, 2012). As these instruments assess observable work-related risk factors, they provide a more tangible starting point for the development of job design solutions (e.g. task interruption or feedback systems, Frese and Zapf, 1999). However, such instruments are likely to suffer from low interrater-reliability due to bias resulting from observers’ individual characteristics (Rugulies, 2012; Semmer et al., 2003). This can be improved by introducing a training curriculum, which trains observers to become familiar with the instrument as well as the underlying constructs. Furthermore, observation-based and self-rated job analysis instruments examine and evaluate work-related risk factors from different perspectives (Frese and Zapf, 1999; Semmer et al., 2003). As a consequence, empirical findings comparing both kinds of measurement tools point to high associations for some psychosocial factors (e.g. task variety, participation, control) and to low associations for job demands,
198
199
Skilled care employees, skilled care assistants
Jobs in a white goods company: assembly operations, picking, material testing Tax officers, executive officers, managers
See 2)
Jobs in a steel mill and a woodworking company: material testing, jobs in the high-bay storage, mechanics, electricians, metal workers, steering system, finishing line, carpenters, engineers, administration See 2)
Jobs in an aircraft factory and several administration units: assembly operations, production planning, controlling, office, library, administration, software development, project coordination, leadership Skilled care employees, skilled care assistants, unskilled employees, employees in non-health care professions (e.g. administration, kitchen, facility management)
Jobs observed
Note. aSamples used for correlational analyses. bSamples used for investigation of criterion validity. ♀ = female. ♂ = male.
8b
Blue collar jobs White collar jobs (public service) Health care Total
Health care
4a,b
6b 7b
Blue collar jobs
3b
Health care
Health care
2b
5a,b
White and blue collar jobs
1b
Setting
Table 1 Research settings (characteristics of collected subsamples, the number and type of SGA profiles, the number of subjective questionnaires).
48 641
59 81
85
96
95
33
144
N SGA profiles
–
– 453
– –
85
< 20: 4.2% 21–30: 32.3% 31–40: 7.3% 41–50: 17.7% 51–60: 21.9% > 60: 1.0% < 20: 9.4% 21–30: 17.6% 31–40: 16.5% 41–50: 27.1% 51–60: 25.9% > 60: 3.5% – –
–
– –
–
♀: 16 ♂: 78 ♀: 65 ♂: -
♀: 31 ♂: 2
–
– < 20: 21–30: 9.1% 31–40: 27.3% 41–50: 30.3% 51–60: 21.2% > 60: –
Sex
Age
96
95
33
144
N matched questionnaire data
A. Tomaschek et al.
Safety Science 101 (2018) 197–208
Safety Science 101 (2018) 197–208
A. Tomaschek et al.
2. Methods
work-related factors do not arise during the observation period or are not directly observable, the observer additionally interviews the employee. In that case, participants answer the standardized items of the SGA instrument normally used for observation. Observers are instructed to keep the interview questions at a minimum. For all steps of the job analysis the SGA contains a manual with detailed instructions in order to guarantee a high quality of assessment. Additionally, it is recommended that the analysis is performed by trained observers. Therefore, a standardized, time-efficient (lasting one to two days depending on previous knowledge) training curriculum for the application of the SGA has been developed during the validation process (Buruck and Richter, 2008). A subgroup of job analyses (N = 28 pairs of SGA profiles) was used to calculate interrater reliability for the psychosocial items (kappa statistic for categorical data, Cohen, 1960). Interrater reliability for the items of Work Activity ranged between 75% and 97%, with kappas of 0.41–0.75 (moderate to substantial reliability; Landis and Koch, 1977), for the items of Work Organization between 68% and 97%, with kappas of 0.72–0.93 (substantial to excellent reliability), except for the item Transparency of work process (93%, kappa = −0.04, prevalence problem) and Cooperation (68%, kappa = 0.35) and for the items of Social Conditions between 79% and 82%, with kappas of 0.43–0.69 (moderate to substantial reliability). Because the magnitude of kappa is affected by the prevalence of the rating, additional parameters (the observed and the expected proportion of agreement, the prevalence index, and the bias index) need to be taken into account in order to interpret the magnitude of kappa meaningfully (Byrt et al., 1993). Prevalence problems exist for some SGA items resulting in low kappa values independent of the true reliability of the instrument. The items along with the corresponding kappa coefficients as well as the additional parameters are presented in Appendix A Table A.1. Self-report job analysis instrument. In order to analyse the construct validity of the SGA, the German version of the Short Questionnaire for the Analysis of Work (SQUAW, Kurz-Fragebogen zur Arbeitsanalyse, Prümper, 2010; Prümper et al., 1995) was used as a self-report measure of psychosocial factors at work. The SQUAW is a validated screening instrument assessing eleven central aspects of an incumbents work situation. Four work-related factors are measured: Work Content, Stressors, Resources and Organization. Work Content is represented by the subscales Versatility (learning of new things, use of knowledge and expertise, task variety) and Completeness (evaluation of the work on the final result, completeness of work product). Stressors are assessed in terms of Qualitative Stress at Work (complexity of work, requirement for concentration), Quantitative Stress at Work (time pressure, workload), Work Interruptions (availability of required information, interruptions by other people) and Environmental Stress (unfavourable environmental factors, inadequate rooms and equipment). Resources refer to Job Latitude (self-determination of working steps, influence on assigned work, independent planning and arrangement of work), Social Backing (reliance on colleagues, reliance on supervisor, cohesion in the department) and Cooperation (requirement of close cooperation, possibility of exchange with colleagues, feedback from supervisors and colleagues). Finally, the factor Organization is covered by the subscales Information and Participation (information about important events, consideration of ideas and suggestions), as well as Company Benefits (training opportunities, career development possibilities). Each subscale consists of two or three items (in total 26), which are answered on a 5-point frequency scale ranging from 1 (very little or strongly disagree) to 5 (very much or strongly agree). Items for each subscale are summed up. High scores indicate well-designed workplaces. In the current study, Cronbach’s α was acceptable to good (α = 0.60–0.85) with the exception of Cooperation (α = 0.43), Work Interruptions
2.1. Sample and procedure The sample for the current study was drawn between 2006 and 2012 from a variety of work settings in Germany: health care (4 samples), blue collar jobs (3 samples), and white collar jobs (2 samples). For data acquisition, observers visited participants at their workplace to conduct an observational interview using the SGA instrument during regular working hours. The assessment lasted on average one hour per job. The participants were asked to carry out their usual work routine while the observation was run. The total sample comprised 641 SGA profiles of various jobs, e.g. secretary, administrator, construction worker, nurse, cook, facility management, type setter. 12 profiles had to be excluded from further analyses due to a large number of missing data. In five samples, subjective questionnaire data from incumbents holding the observed jobs were available. Information concerning study settings, the number of SGA profiles, the type of screened jobs, and (if available) the number of subjective questionnaire data (for a description see Section 2.2) with corresponding demographic information is listed in Table 1. 2.2. Materials Observation-based job analysis instrument. The “Healthy Workplace Screening” (Screening Gesundes Arbeiten, SGA – Debitz et al., 2014; Richter et al., 2014) is a standardized job analysis instrument measuring work-related risk factors on a screening level by combining behavioural observation with a complementary interview technique. The SGA was developed as an adapted low-threshold and user-friendly version of the job analyses instrument “REBA” (Rechnergestützte psychologische Bewertung von Arbeitstätigkeiten - for more details on the development procedure see e.g. Richter et al., 2009). In contrast to selfreport questionnaires, where dimensions are often obtained by classical test theory techniques (e.g., factor analysis), dimensions of this observation–based instrument were aggregated by theoretical considerations. Hence, the items were derived from state-of-the art job design models: Hacker’s model of action regulation (Hacker, 2001), the JobDemand-Control-Support Model (Johnson and Hall, 1988) and the Effort-Reward-Imbalance Model (Siegrist, 1996; Siegrist et al., 2004). In addition, the “Cinderella” - Hypothesis (Hägg, 1991, 2000; Melin and Lundberg, 1997) was used to apply a more integrated approach to the assessed work-related risk factors, namely including physical and psychosocial factors. The instrument measures 44 items grouped into eight dimensions, representing a combination of meaningful physical and psychosocial work-related risk factors as well as environmental conditions (for an overview of the SGA dimensions and example items see Appendix Table A1). The items allow the direct development of intervention suggestions for job design providing specific information how to improve the certain work-related factor. The physical factors investigated by the SGA represent three dimensions: Posture (3 items), Anthropometry (5 items) and Tools & Devices (3 items). The analysed psychosocial factors comprise three dimensions as well: Work Activity (9 items), Work Organization (5 items) and Social Conditions (3 items). Environmental conditions are further divided into the following two dimensions: Work Contract (7 items) and Work Environment (4 items). In this study environmental conditions are included mainly for the practical reason of drawing implications for job design after analysis. For each item, the observer states whether or not the work-related factor represents a risk (dichotomic ratings, yes vs. no). At the end, a sum score is created for each dimension, adding up the number of items categorized as a “risk”. If
200
Safety Science 101 (2018) 197–208
A. Tomaschek et al.
3.2. Construct validity
(α = 0.48), Completeness (α = 0.53), Qualitative Stress at Work (α = 0.58) and Versatility (α = 0.59). Measure of physical health. In order to examine criterion validity of the SGA, the Freiburg Complaint List (Freiburger Beschwerdeliste, FBLR Fahrenberg, 1995) was applied. This established standardized questionnaire measures current (i.e., situationally caused) as well as chronic health impairments. For the present study the subscale referring to musculoskeletal complaints (covering 8 items, including back, shoulder, neck, and leg pain) was used. The participants rated the frequency of their symptoms on a 5-point scale ranging from 1 (never) to 5 (almost every day). In the current study, Cronbach’s α was good (α = 0.82). Measure of mental health. As another indicator of criterion validity, mental health was assessed using a shorted version of the General Health Questionnaire (GHQ-12, Goldberg and Williams, 1988). The Questionnaire consists of 12 items (e.g., “Over the past weeks, have you felt capable of making decisions about things?” or “Over the past weeks, have you lost much sleep because of worry?”). Each item evaluates the severity of a mood state over the past 4–6 weeks on a 4-point Likertscale (from 0 to 3). The GHQ-12 score is computed as the sum of all items. High scores are interpreted as poor mental health. In the current study, Cronbach's alpha calculated for this scale was good (α = 0.85).
Regarding convergent validity, small to medium correlations ranging from r = -0.14 up to r = -0.35 between the SGA items and the subjective SQUAW subscales (all significant on p < 0.010, see Table 2) were found. Especially for the SGA items indicating Participation, Leadership, Reward, Repetition and Responsibility, the analyses showed the expected associations with the corresponding subjective SQUAW subscales. 3.3. Criterion validity First, to determine job groups with similar SGA risk patterns, a cluster analysis was performed. The analysis yielded a 3-cluster-solution. According to the plot of agglomeration coefficients and the dendrogram (see Appendix, Fig. A.3), the three clusters explained the data most adequately and were represented by a comparable number of observations (N1 = 198, N2 = 190, N3 = 180). Table 3 presents the means, standard deviations and results of univariate tests of the SGA work dimensions for each of the three clusters. Cluster one showed the lowest risk levels in all SGA dimensions except for Posture. Cluster two was characterized by the highest risk in the physical factors (Posture, Tools & Devices and Anthropometry), whereas cluster three had the highest risk levels in psychological factors (Work Activity, Work Organization, Social Conditions).2 Post-hoc comparisons showed that all clusters differed significantly for most SGA dimensions at p < 0.05. Cluster one and three did not differ significantly in regard to Tools & Devices and Anthropometry and cluster two and three in regard to Social Conditions. Next, the predictive value of the clusters with regard to physical and psychological outcome variables was examined. Univariate ANOVAs resulted in significant differences between the clusters with regard to mental health, overall musculoskeletal complaints and back pain. Posthoc contrasts revealed that individuals in the low-risk-cluster report significant higher subjective mental health as well as lower levels of musculoskeletal complaints and back pain than individuals in the highrisk clusters (clusters 2 and 3), which are characterized by predominantly physical and psychosocial risk factors, respectively. The latter two did not differ significantly from one another. Mean scores and standard deviations of the outcome variables for each cluster as well as results of the ANOVAs are shown in Table 4.
2.3. Statistical procedure To answer the question about the validity of the SGA, statistical tests to define construct as well as criterion validity were performed. All statistical analyses were conducted using IBM SPSS statistics package (Version 23.0). Subjective data was matched to SGA profile by surveying the specific job holder, i.e. after observation with the SGA, incumbents were asked to fill in a questionnaire. As an aspect of construct validity, it was investigated, how psychosocial factors measured by the SGA items relate to the corresponding SQUAW subscales. To estimate the degree of relationship between the dichotomous SGA items and the continuous SQUAW scales the point-biserial correlation coefficient was used. Mathematically, the point-biserial correlation is equivalent to the Pearson’s product-moment correlation. Since the SGA instrument assesses actual risks whereas high scores in the SQUAW indicate well-designed work - high negative correlations between both measures suggest a good validity. To investigate criterion validity, job groups with similar risk patterns in the SGA profiles were identified by a hierarchical cluster analysis. Specifically, the Ward's Method was applied, an agglomerative procedure that maximizes the heterogeneity between the clusters compared to within-cluster-variance. As a distance metric squared Euclidean distance was used. Before initiating the cluster analysis, the data was z-standardized. After this procedure, one-way ANOVAs along with Scheffé-Tests for Post-Hoc comparisons were applied to compare the identified SGA clusters in regard to relevant health-related outcome variables, namely poor mental health, overall musculoskeletal complaints and back pain.
4. Discussion The present study aimed at the validation of the “Healthy Workplace Screening“, an observation-based screening instrument developed for the analysis, evaluation and design of work-related risk factors in a variety of branches (SGA, Buruck et al., 2007; Debitz et al., 2014). So far, user-friendly, feasible and time-efficient observationbased measures are rare (Gebele et al., 2011; Grebner et al., 2005; Leitner and Resch, 2005; Rugulies, 2012). Aiming at bridging this gap, the SGA instrument allows for a first estimation of physical and psychosocial risk factors on the basis of an observational interview (combination of an observation of the work activity and an oral interview with the job incumbent). Overall results regarding construct validity showed that the jobrelated information measured by SGA items are similar to the information covered by the stressor subscales of the validated SQUAW questionnaire, supporting convergent validity. Second, concerning criterion validity, the ANOVAs comparing statistically identified clusters
3. Results 3.1. Descriptive results The analysis revealed small, medium and large inter-scale correlations between several SGA dimensions. The correlations along with means and standard deviations for each SGA dimension are presented in the Appendix Table A.2.
2 A 2 - cluster solution separated observations into a high and a low risk group whereas cluster 2 and cluster 3 in the 3 – cluster solution were merged in a single cluster.
E-mail addresses:
[email protected] (A. Tomaschek),
[email protected] (G. Buruck).
201
202
Work cycle Responsibility Cooperation Stability of cooperation Participation Short rest breaks
Social Support Leadership Reward
SGA work organization
SGA social conditions
0.10 −0.22** −0.07
0.01 0.05
0.04 0.05 0.03 /
−0.07 0.00 −0.05
0.05 0.01 0.07 0.04 0.05 0.03
SQUAW Completeness
0.13 −0.02 −0.12 −0.02 −0.02
−0.03 0.04 0.01
0.05 0.09 −0.01 /
−0.02 0.00 −0.03
−0.30** 0.17* 0.18* −0.26** −0.30** 0.14+
SQUAW Cooperation
−0.05 0.02
0.03 0.25** 0.02 /
−0.05 −0.13+ −0.07
−0.09 −0.13+ 0.12 0.04 −0.09 0.01
SQUAW Social backing
−0.26** −0.03 0.05
−0.11 0.09
0.03 0.05 −0.14+ /
0.00 −0.03 −0.01
−0.35** 0.05 0.00 −0.31** −0.35** 0.07
SQUAW Qualitative stress at work
p < 0.01. /could not be computed in SPSS because one variable is constant.
**
0.00 −0.28** −0.16*
−0.06 −0.07
−0.05 −0.15+
0.00 −0.17* −0.20*
0.11 0.00 −0.04 /
−0.22** −0.04 −0.14+
−0.12 −0.09 0.03
0.08 −0.27** 0.09 /
−0.12 −0.01 0.01 −0.15+ −0.12 0.02
−0.03 0.10 −0.15+ −0.08 −0.03 0.21**
Note. N = 162–164.+p < 0.10, *p < 0.05,
Task Variety Job Demands Task Identity Repetition Control Work task conflicts Feedback Information Customer interaction
SGA work activity
SQUAW Versatility
SQUAW Job latitude
Table 2 Intercorrelations among SGA and SQUAW work-related factors.
−0.05 −0.02 −0.08
−0.24** −0.04 −0.09
−0.08 0.01
0.03 0.21** −0.08 /
−0.04 0.05 0.16* / −0.04 0.02
−0.05 0.03 −0.04
−0.20* −0.04 0.10 −0.19* −0.20* 0.05
SQUAW Work interruptions
0.04 0.01 0.06
0.17* −0.12 0.03 0.11 0.17* 0.05
SQUAW Quantitative stress at work
−0.29** −0.14+ −0.02
−0.17* 0.06
0.09 0.15+ −0.12 /
−0.03 0.01 −0.08
−0.32** −0.07 0.05 −0.27** −0.32** 0.10
SQUAW Environmental stress
0.02 −0.22** −0.32**
−0.07 0.14+
−0.18* −0.06 −0.06 −0.27** −0.23**
0.09 0.05 −0.03 /
−0.13+ −0.12 0.00
−0.21** 0.13 −0.06 −0.20* −0.21** 0.22**
SQUAW Company benefits
0.08 0.06 0.09 /
−−0.20** −0.10 −0.06
−0.14+ −0.02 0.08 −0.19* −0.14+ 0.11
SQUAW Information & participation
A. Tomaschek et al.
Safety Science 101 (2018) 197–208
Safety Science 101 (2018) 197–208
A. Tomaschek et al.
Table 3 Cluster differences in SGA work-related factors. Cluster 1a
Posture Tools and devices Anthropometry Work activity Work organization Social conditions
Cluster 2a
Cluster 3a
M
SD
M
SD
M
SD
1.33 0.36 0.22 1.83 0.72 1.04
0.71 0.56 0.44 1.11 0.93 0.76
1.97 1.58 1.74 3.03 1.09 1.11
1.06 0.93 0.95 1.52 1.00 0.77
0.86 0.42 1.63 3.99 2.46 1.23
0.64 0.54 0.66 0.86 1.12 0.63
F(2547–567)
Post Hoc
83.53*** 181.46*** 274.07*** 154.20*** 149.95*** 3.18*
[1,2][1,3][2,3] [1,2][2,3] [1,2][1,3] [1,2][1,3][2,3] [1,2][1,3][2,3] [1,3]
Note. N = 548–568.aBased on zscores. Post hoc test (Scheffé): denoting pairs of clusters that are significantly different at the 0.05 level.
subjective health-related outcomes (poor mental health and musculoskeletal complaints). Cluster analyses identified three meaningful groups: A first group indicating well-designed work characterized by the lowest risk regarding the work-related factors measured by the SGA (low risk cluster) and two groups representing the highest risk regarding the physical and psychosocial factors measured by the SGA, respectively (high risk cluster). Analysis revealed that the low risk cluster significantly differs from the two high risk clusters with regard to perceived mental health and musculoskeletal complaints, especially back pain. Whereas the high risk cluster with predominantly physical demands did not differ significantly from the high risk cluster representing mainly psychosocial demands. Thus, employees with jobs characterized by lower risks in work-related factors measured by the SGA instrument reported better mental health and lower back pain than employees with jobs characterized by higher physical and higher psychosocial risk factors. These results are in line with the state of the art of occupational health research, which provides evidence for a linkage between negative work-related risk factors and impaired mental as well as somatic health. Specifically, several meta-analyses and reviews based on cross-sectional as well as longitudinal data confirm the effects of psychosocial factors - especially of work load, autonomy, social support and effort/reward imbalance - on mental health (de Lange et al., 2003; Harvey et al., 2017; Häusser et al., 2010; Rau and Buyken, 2015; Theorell et al., 2015; Tsutsumi and Kawakami, 2004) and musculoskeletal complaints (Hauke et al., 2011; Lang et al., 2012; Larsman et al., 2011). In addition to psychosocial factors, there is also extensive empirical evidence for health risks due to physical stressors (Ariens et al., 2000; Burr et al., 2017; Gerr et al., 2014; Hoogendoorn et al., 2002). The present results do not suggest any differences in mental health and back pain between the two high-risk clusters with predominantly psychosocial and physical work risk factors, respectively. This finding supports the “Cinderella” - Hypothesis (Hägg, 1991, 2000; Melin and Lundberg, 1997) and several empirical studies (Kouvonen et al., 2016; Lundberg et al., 2002; Nixon et al., 2011; Sjøgaard et al., 2000; Taylor et al., 2014) underlining the necessity to monitor both psychosocial as well as physical factors for the estimation of harmful health outcomes. Thus, the results support criterion validity of the SGA instrument. Unexpected results were found for the SGA dimension Posture: With regard to the examination of criterion validity, jobs characterized by low risks in the SGA dimensions Anthropometry, Tools & Devices, Work Activity, Work Organization, and Social Conditions simultaneously show higher risks in the (physical) SGA dimension Posture. This might be due to the fact that most of the jobs require either extensive standing (e.g. in blue collar jobs), extensive sitting (e.g. in white collar jobs) or both demands (e.g. in health care jobs). Thus, for this SGA item the observer marks a risk (representing either sitting, standing or both) for almost any job and is not able to distinguish between different job types
Table 4 Cluster differences in mental health, musculoskeletal complaints and back pain.
Mental health Musculoskeletal complaints Back Pain
Cluster 1a
Cluster 2a
Cluster 3a
M
SD
M
SD
M
SD
7.29
4.93
8.98
5.82
9.04
F(2358–404)
Post Hoc
5.45
4.866** **
[1,2] [1,3] [1,2] [1,3] [1,2] [1,3]
16.21
6.86
18.41
7.11
19.26
7.25
6.848
2.57
1.42
3.09
1.31
3.18
1.29
8.466***
Note. N = 359–405. aBased on z scores. Post hoc test (Scheffé): denoting pairs of clusters that are significantly different at the 0.05 level.
of distinguishable SGA profiles demonstrated that jobs with varying risks in SGA dimensions significantly differ with regard to mental health and musculoskeletal complaints. 4.1. Construct validity With regard to convergent validity, results revealed small to medium-sized correlations between the observer-based SGA measure and the self-report SQUAW instrument. As an explanation for the rather weak to moderate associations, a method effect due to the divergence of observation-based and self-report assessment should be considered. For instance, it can be argued that some subscales of the SQUAW (such as Versatility or Completeness) rather focus on the workers appraisal than on the objective assessment of risk factors in the work environment per se. In line with this and with previous empirical findings comparing self-report and observation-based measures (Demerouti, 1999; Nachreiner, 2008; Rau et al., 2010; Waldenström and Harenstam, 2008), higher correlations were found for SGA risk factors that can be easily observed (e.g. Task Variety, Participation, Control). In contrast, Job demands, Responsibility, Cooperation and Conflicts were weakly related. In addition, the SGA items were particularly associated with the Stressor subscales of the SQUAW instrument (Qualitative and Quantitative Stress at Work, Work Interruptions, and Environmental Stress). This might be explained by the fact that both the SQUAW Stressor subscales and the SGA items strive on assessing risk factors at work. In contrast, the SQUAW subscales representing Resources and Organization did not relate to most of the SGA items. 4.2. Criterion validity Criterion validity was assessed by investigating associations between job groups of similar SGA work-related risk factors and
203
Safety Science 101 (2018) 197–208
A. Tomaschek et al.
knowledge) standardized training curriculum for practitioners is available (Buruck and Richter, 2008). It ensures reliable assessments by knowledge transfer and case studies.
per se. Therefore, it is recommended to interpret the single items instead of the whole Posture dimension. 4.3. Limitations
4.4. Implications and conclusion With regard to the current study, several limitations need to be considered. First, it can be argued that the generalizability of the results is limited to the collected samples, i.e. job profiles. Four of the eight data sets were obtained in the field of health care. However, all health care data sets also contained non-care jobs (i.e. administration, facility management) and a total of 379 out of 641 SGA profiles were collected in blue and white collar settings. Therefore it can be argued that the SGA profiles used in the current study represent a wide range of jobs and occupational groups. Another limitation of the present study concerns the cross-sectional design. Regarding construct validity, cross-sectional data are sufficient for analysing convergent validity with subjective measures. However, with regard to criterion validity, a cross-sectional design does not allow the causal prediction of occupational health outcomes. However, in line with previous research (e.g., Lang et al., 2012; Rau and Henkel, 2013), the current results showed that employees occupied in jobs with low risks regarding the SGA work-related factors report better mental health and less back pain compared to employees with jobs characterized by higher work-related risk factors. Nevertheless, further studies should additionally analyse the effects of jobs with high-risk SGA profiles on occupational health outcomes in the long term. In addition, the convergent validity of the SGA was only assessed with a self-report instrument measuring psychosocial factors (SQUAW). This instrument was selected for investigating the convergent validity in our sample of health care workers (especially nurses) since this measure allows for the evaluation of work-related factors in the (health) care setting (Ehlbeck et al., 2008). Furthermore, the SQUAW, similar to the SGA instrument, is also used for conducting risk analyses at work (“Gefährdungsanalysen”) based on the German Safety and Health at Work Act (“Arbeitsschutzgesetz”). However, it might be argued that other observation-based and/or subjective measures are needed to ensure valid results, particularly for the physical factors. In our samples the application of another observation-based tool was not possible since observing one workplace with two observers at the same time is very challenging. For further studies we would strongly recommend to address this issue. Likewise, it should be taken into account, that the assessment of mental and physical health were based on self-reports. Although workers are the obvious experts of their own health situation and standardized valid instruments were applied, future studies should be complemented by more objective criterion measures like e.g. archival (e.g. sickness leave, performance measures, accidents) or biological data. (e.g. cortisol). Finally, for a subsample of profiles which were assessed in pairs, the interrater reliability was calculated. For some work-related factors - e.g. factors mandatorily necessitating specific observational knowledge (Work Space, Unnatural Posture) or factors difficult to observe (Leadership, Social Support) the interrater reliability was low. As dual observations are usually performed at the beginning of a study in order to familiarize the observers with the instrument and to discuss disagreements concerning e.g. the items, the results emphasize the importance of training users for observation-based assessments. At present, a time-efficient (lasting one to two days depending on previous
The design of healthy workplaces requires the identification (and – if necessary – reduction) of potential work-related risk factors. For this purpose, valid comprehensive and efficient instruments are needed. Based on the results of the current study, the application of the observation-based “Healthy Workplace Screening” (SGA) as a suitable screening instrument for the analysis of jobs is proposed. The SGA integrates both psychosocial and physical risk factors in the work environment, which are predictive for musculoskeletal pain and impaired mental health, and allows for distinctions between occupations characterized by different demands as well as between high and low risk jobs. Thus, the SGA is an important complement to widely-used subjective assessments in order to determine different perspectives of work-related risk factors. To our knowledge, valid as well as userfriendly observation-based instruments are currently limited. Compared to self-report measures the SGA has the advantage that ethical considerations, like data privacy questions, play a minor role since the focus of observation-based assessment is on workplace characteristics, not on the rating of a certain job holder. Nevertheless, to ensure confidentiality of data it is recommended to apply SGA for jobs with at least five incumbents holding the same position. In sum, the SGA builds a bridge between scientific and practitioners’ requirements by promoting a user-friendly, integrative workplace risk assessment. Moreover, the standardized observer training curriculum as well as the design advices and solutions included in the instrument make the SGA a low-threshold tool for workplace health promotion supporting healthy workplaces, and thus, healthy employees. Acknowledgment First of all, we are grateful to Peter Richter, who initiated the process of SGA development. Our special thanks also go to Hildegard Schmidt, who died suddenly last year. With great commitment, she introduced the SGA in the important German health and safety institutions and initiatives. Furthermore, we thank our colleagues Susann Mühlpfordt, Elke Muzykorska and Ulrike Lübbert as well as the former master students Jens Harloff, Cindy Eibisch, Irén Horváth, and Sina Keller for their important support in collecting the data. Our gratitude also goes to the participating companies that supported the conduction of job analyses; without those, the present study would not have been realised. Finally, we thank Denise Dörfel and Ina Zwingmann for their important advice on this paper. Funding This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. Disclosure Statement The authors declare that there are no conflicts of interest.
Appendix A Tables A.1 and A.2 and Fig. A.3.
204
Safety Science 101 (2018) 197–208
A. Tomaschek et al.
Table A.1 SGA dimensions with corresponding items and kappa coefficients, the observed and the expected proportion of agreement, the prevalence, and the bias index. Item Work contract
Posture
Tools and devices
AV1 AV2 AV3 AV4 AV5 AV6 AV7
Work contract Temporary employment Free-lance employment Permanent workplace Night shifts On-call duty Overtime
pH1
Sitting
pH2 pH3 pH4 pH5
Ergonomical sitting Standing Ergonomical standing Moving
pH6
pH9
Ergonomical use of transportation devices Availability of work equipment Tools supporting healthy body posture Functionality of work equipment
pH10
Vibration load
pH11 pH12
Load movement Repetitive movements
pH13 pH14 PH15
Work Space/operating range Reach distance Work height
PS1 PS2 PS3
Task Variety Job demands Task completeness
pH7 pH8
Anthropometry
Work activity
Work organization
Social conditions
Work environment
Kappa
Po
Pe
Prevalence
Bias
Example
Does the work contract include night-shifts on a regular basis?
Does the work task require a sitting position over a period longer than 20 minutes?
Does the work equipment function properly to support an undisturbed execution of the work task activities?
Does the work task require to perform repetitive movements for a period longer than 20 minutes?
0.75*** 0.65*** 0.41*
89 97 75
58 90 58
0.39 0.89 0.39
0.04 0.04 0.04
***
PS4 PS5 PS6 PS7 PS8 PS9
Task repetitiveness Work scheduling autonomy Conflicts in task requirements Feedback Information Customer interaction
0.59 0.72*** / 0.44* 0.47* /
86 86 / 75 93 /
65 49 / 55 83 /
0.57 0.00 / 0.32 0.85 /
0.14 0.14 / 0.04 0.07 /
PS10 PS11
Transparency of work process Decision-Making Autonomy
-0.04 0.72***
93 86
93 49
0.93 0.07
0.00 0.14
PS12 PS13 PS14 PS15
Cooperation Stability of cooperation relations Participation Short rest breaks
0.35 0.93*** 0.75*** 0.78***
68 97 89 89
51 51 58 51
0.11 0.11 0.39 0.18
0.04 0.04 0.07 0.04
PS16
Social Support
0.43*
79
62
0.50
0.07
PS17 PS18
Leadership Reward
0.69*** 0.61***
80 82
54 54
0.29 0.32
0.00 0.11
AU1 AU2 AU3 AU4
Lighting Noise Smell Room climate
Does the work task include preparation, coordination or reviewing activities?
Do employees have the possibility to make their own decisions concerning the work task?
Is support from other employees available when work task related problems occur?
Is there adequate light for the work task?
Table A.2 Means, SDs and inter-scale correlations for SGA dimensions. Dimension
N
M
SD
Variables indicated by numbers 2
1. 2. 3. 4. 5. 6. 7. 8.
Posture (5) Tools and devices (5) Anthropometry (5) Work activity (9) Work organization (6) Social conditions (3) Work contract (7) Work environment (4)
607 621 624 628 629 622 602 624
1.40 0.76 1.14 2.63 1.38 1.08 1.05 1.57
0.93 0.88 0.98 1.56 1.22 0.74 1.04 1.11
0.31
3 **
0.04 0.15**
205
4
5
0.00 0.07* 0.35**
−0.11 −0.03 0.21** 0.48**
6 **
−0.16 0.06 0.19** 0.16** 0.23**
**
7
8
0.04 0.10* 0.45** 0.13** 0.08* 0.12**
−0.07 −0.07 0.34** 0.29** 0.28** 0.18** 0.12**
Safety Science 101 (2018) 197–208
A. Tomaschek et al.
Fig. A.3. Dendrogram of cluster analysis.
206
Safety Science 101 (2018) 197–208
A. Tomaschek et al.
Appendix B. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.ssci.2017.09.006.
und Decision Latitude sowie Zusammenhänge der Tätigkeitsmerkmale mit Erholungsunfähigkeit [Objective determination of job demands and decision latitude as well as links of job characteristics with inability of recovery]. Zeitschrift für Arbeits- und Organisationspsychologie 55, 32–45. http://dx.doi.org/10.1026/09324089/a000036. Gerr, F., Fethke, N.B., Merlino, L., Anton, D., Rosecrance, J., Jones, M.P., Marcus, M., Meyers, A.R., 2014. A prospective study of musculoskeletal outcomes among manufacturing workers: I. effects of physical risk factors. Hum. Fact.: J. Hum. Fact. Ergon. Soc. 56, 112–130. http://dx.doi.org/10.1177/0018720813491114. Goldberg, D., Williams, P., 1988. A User’s Guide to the General Health Questionnaire. NFER-Nelson, Windsor, UK. Grebner, S., Semmer, N.K., Elfering, A., 2005. Working conditions and three types of wellbeing: a longitudinal study with self-report and rating data. J. Occup. Health Psychol. 10, 31–43. http://dx.doi.org/10.1037/1076-8998.10.1.31. Hacker, W., 2001. Activity theory: psychological. In: Smelser, N.J., Baltes, P.B. (Eds.), International Encyclopedia of the Social & Behavioral Sciences. Pergamon, Oxford, pp. 58–62. Hacker, W., Fritsche, B., Richter, P., Iwanowa, A., 1995. Das Tätigkeitsbewertungsverfahren TBS-L [A Method for Job Analysis, Job Evaluation, and Job Design]. Verlag der Fachbuchvereine, Zürich. Hägg, G.M., 1991. Static work loads and occupational myalgia - a new explanation model. In: Andersson, P.A., Hobart, D.J., Danoff, J.V. (Eds.), Electromyographical Kinesiology. Elsevier Science Publishers, Amsterdam, pp. 141–144. Hägg, G.M., 2000. Muscle fibre abnormalities related to occupational load - a review. Eur. J. Appl. Physiol. 83, 159–165. Harvey, S.B., Modini, M., Joyce, S., Milligan-Saville, J.S., Tan, L., Mykletun, A., Bryant, R.A., Christensen, H., Mitchell, P.B., 2017. Can work make you mentally ill? A systematic meta-review of work-related risk factors for common mental health problems. Occup. Environ. Med. 74, 301–310. http://dx.doi.org/10.1136/oemed-2016104015. Hauke, A., Flintrop, J., Brun, E., Rugulies, R., 2011. The impact of work-related psychosocial stressors on the onset of musculoskeletal disorders in specific body regions: a review and meta-analysis of 54 longitudinal studies. Work Stress 25, 243–256. http://dx.doi.org/10.1080/02678373.2011.614069. Häusser, J.A., Mojzisch, A., Niesel, M., Schulz-Hardt, S., 2010. Ten years on: a review of recent research on the Job Demand-Control (-Support) model and psychological wellbeing. Work Stress 24, 1–35. http://dx.doi.org/10.1080/02678371003683747. Hoogendoorn, W.E., Bongers, P.M., de Vet, H.C.W., Ariëns, G.A.M., van Mechelen, W., Bouter, L.M., 2002. High physical work load and low job satisfaction increase the risk of sickness absence due to low back pain: results of a prospective cohort study. Occup. Environ. Med. 59, 323–328. http://dx.doi.org/10.1136/oem.59.5.323. Horváth, I., Buruck, G., Richter, R., 2015. Der Einfluss von Arbeitsbedingungen auf Rückenschmerzen [Impact of working conditions on back pain]. Pflegewissenschaft 17, 36–45. http://dx.doi.org/10.3936/1284. Johnson, J.V., Hall, E.M., 1988. Job strain, work place social support, and cardiovascular disease: a cross-sectional study of a random sample of the Swedish working population. Am. J. Public Health 78, 1336–1342. Keller, S., Buruck, G., Debitz, U., Richter, R., 2010. Identifizierung tätigkeitsspezifischer Belastungsmuster am Arbeitsplatz [Identification of activity-specific load patterns at work]. In: Rigotti, T., Korek, S., Otto, K. (Eds.), Gesund Mit Und Ohne Arbeit. Pabst Science Publishers, Lengerich, pp. 203–218. Kouvonen, A., Manty, M., Lallukka, T., Lahelma, E., Rahkonen, O., 2016. Changes in psychosocial and physical working conditions and common mental disorders. Eur. J. Pub. Health 26, 458–463. http://dx.doi.org/10.1093/eurpub/ckw019. Kuhn, K., 2007. Arbeitsbedingte Einflüsse bei der Entstehung chronischer Krankheiten [Work-related impacts of developing chronic diseases]. In: Badura, B., Schellschmidt, H., Vetter, C. (Eds.), Fehlzeiten-Report 2006. Springer, Berlin Heidelberg, pp. 25–43. Landis, J.R., Koch, G.G., 1977. The measurement of observer agreement for categorical data. Biometrics 33, 159–174. Lang, J., Ochsmann, E., Kraus, T., Lang, J.W.B., 2012. Psychosocial work stressors as antecedents of musculoskeletal problems: a systematic review and meta-analysis of stability-adjusted longitudinal studies. Soc. Sci. Med. 75, 1163–1174. http://dx.doi. org/10.1016/j.socscimed.2012.04.015. Larsman, P., Lindegard, A., Ahlborg, G., 2011. Longitudinal relations between psychosocial work environment, stress and the development of musculoskeletal pain. Stress Health 27, E228–E237. http://dx.doi.org/10.1002/smi.1372. Leitner, K., Resch, M.G., 2005. Do the effects of job stressors on health persist over time? A longitudinal study with observational stressor measures. J. Occup. Health Psychol. 10, 18–30. http://dx.doi.org/10.1037/1076-8998.10.1.18. Lundberg, U., 2002. Psychophysiology of work: stress, gender, endocrine response, and work-related upper extremity disorders. Am. J. Ind. Med. 41, 383–392. http://dx.doi. org/10.1002/ajim.10038. Lundberg, U., Forsman, M., Zachau, G., Eklöf, M., Palmerud, G., M, B., Kadefors, R., 2002. Effects of experimentally induced mental and physical stress on trapezius motor unit recruitment. Work Stress 16, 166–178. http://dx.doi.org/10.1080/ 02678370210136699. Melin, B., Lundberg, U., 1997. A biopsychosocial approach to work-stress and musculosketal disorders. J. Psychophysiol. 11, 238–247. Milner, A., Krnjacki, L., LaMontagne, A.D., 2017. Psychosocial job quality and mental
References Ahola, K., Honkonen, T., Kivimaki, M., Virtanen, M., Isometsa, E., Aromaa, A., Lonnqvist, J., 2006. Contribution of burnout to the association between job strain and depression: the health 2000 study. J. Occup. Environ. Med. 48, 1023–1030. http://dx.doi. org/10.1097/01.jom.0000237437.84513.92. Ariens, G.A., van Mechelen, W., Bongers, P.M., Bouter, L.M., van der Wal, G., 2000. Physical risk factors for neck pain. Scand. J. Work Environ. Health 26, 7–19. http:// dx.doi.org/10.5271/sjweh.504. Bernal, D., Campos-Serna, J., Tobias, A., Vargas-Prada, S., Benavides, F.G., Serra, C., 2015. Work-related psychosocial risk factors and musculoskeletal disorders in hospital nurses and nursing aides: a systematic review and meta-analysis. Int. J. Nurs. Stud. 52, 635–648. http://dx.doi.org/10.1016/j.ijnurstu.2014.11.003. Brom, S.S., Buruck, G., Horváth, I., Richter, P., Leiter, M.P., 2015. Areas of worklife as predictors of occupational health – a validation study in two German samples. Burnout Res. 2, 60–70. http://dx.doi.org/10.1016/j.burn.2015.05.001. Burr, H., Pohrt, A., Rugulies, R., Holtermann, A., Hasselhorn, H.M., 2017. Does age modify the association between physical work demands and deterioration of selfrated general health? Scand. J. Work, Environ. Health 43, 241–249, http://dx.doi. org/10.5271/sjweh.3625. Buruck, G., Brom, S.S., Kotte, H., Richter, P., 2015. Evaluierung von Gesundheitspräventionen in der stationären Altenpflege [Evaluating health prevention programms in nursing homes], In: Badura, B., Ducki, A., Schröder, H., Klose, J., Meyer, M. (Eds.), Fehlzeiten-Report 2015: Neue Wege für mehr GesundheitQualitätsstandards für ein zielgruppenspezifisches Gesundheitsmanagement [Absenteeism Report 2015: New Directions for improving Health - Quality standards for a targeted group-specific health management]. Springer-Verlag, Berlin, pp. 85–95. Buruck, G., Debitz, U., Rudolf, M., 2007. Screening Gesundes Arbeiten (SGA) - Erste Ergebnisse einer Pilotstudie [Screening Healthy work (SGA) - First results of a pilot study]. In: Richter, P.G., Rau, R., Mühlpfordt, S. (Eds.), Arbeit Und Gesundheit. Zum Aktuellen Stand in Einem Forschungs- Und Praxisfeld. Pabst Science Publishers, Lengerich, pp. 73–94. Buruck, G., Richter, P., 2008. Psychische Belastungen: Weniger Befragung - mehr Beobachtung [Mental stress: Less survey - more observation]. Gute Arbeit. Gesundheitsschutz und Arbeitsgestaltung 11, 30–32. Byrt, T., Bishop, J., Carlin, J.B., 1993. Bias, prevalence and kappa. J. Clin. Epidemiol. 46, 423–429. Campbell, P., Foster, N.E., Thomas, E., Dunn, K.M., 2013. Prognostic indicators of low back pain in primary care: five-year prospective study. J. Pain 14, 873–883. http:// dx.doi.org/10.1016/j.jpain.2013.03.013. Cohen, J., 1960. A coefficient of agreement for nominal scales. Educ. Psychol. Measur. 20, 37–46. http://dx.doi.org/10.1177/001316446002000104. da Costa, B.R., Vieira, E.R., 2010. Risk factors for work-related musculoskeletal disorders: a systematic review of recent longitudinal studies. Am. J. Ind. Med. 53, 285–323. http://dx.doi.org/10.1002/ajim.20750. Daniels, K., 2006. Rethinking job characteristics in work stress research. Human Rel. 59, 267–290. http://dx.doi.org/10.1177/0018726706064171. de Lange, A.H., Taris, T.W., Kompier, M.A.J., Houtman, I.L.D., Bongers, P.M., 2003. 'The very best of the millennium': longitudinal research and the demand-control-(support) model. J. Occup. Health Psychol. 8, 282–305. http://dx.doi.org/10.1037/1076-8998. 8.4.282. Debitz, U., Buruck, G., Mühlpfordt, S., Muzykorska, E., Lübbert, U., Schmidt, H., 2014. Der Leitfaden zum Screening Gesundes Arbeiten (SGA). Physische und psychische Gefährdungen erkennen - gesünder arbeiten! [The guideline for the Healthy Workplace Screening (SGA). Recognize physical and psychological threats - work healthier!], 3ed. Bundesanstalt für Arbeitsschutz und Arbeitsmedizin, Dortmund. Demerouti, E., 1999. Burnout: Eine Folge Konkreter Arbeitsbedingungen Bei Dienstleistungs- Und Produktionstätigkeiten [Burnout: A Consequence of Working Conditions Among Service Employees and Industrial Workers]. Lang, Frankfurt am Main. Ehlbeck, I., Lohmann, A., Prümper, J., 2008. Arbeit in Krankenhäusern Human Gestalten [Make Working in Hospitals Humane]. Hans-Böckler-Stiftung, Düsseldorf. Esquirol, Y., Niezborala, M., Visentin, M., Leguevel, A., Gonzalez, I., Marquié, J.-C., 2017. Contribution of occupational factors to the incidence and persistence of chronic low back pain among workers: results from the longitudinal VISAT study. Occup. Environ. Med. 74, 243–251. http://dx.doi.org/10.1136/oemed-2015-103443. EU-OSHA, 2012. Management of Psychosocial Risks at Work: An Analysis of the Findings of the European Survey of Enterprises on New and Emerging Risks. European Agency for Safety and Health at Work, Luxembourg. EU-OSHA, 2015. Second European Survey of Enterprises on New and Emerging Risks (ESENER-2). European Agency for Safety and Health at Work, Luxembourg. Fahrenberg, J., 1995. Somatic complaints in the German population. J. Psychosom. Res. 39, 809–817. http://dx.doi.org/10.1016/0022-3999(94)00151-0. Frese, M., Zapf, D., 1999. On the importance of the objective environment in stress and attribution theory. Counterpoint to Perrewe and Zellars. J. Organ. Behav. 20, 761–765. Gebele, N., Morling, K., Rösler, U., Rau, R., 2011. Objektive Erfassung von Job Demands
207
Safety Science 101 (2018) 197–208
A. Tomaschek et al. health among young workers: a fixed-effects regression analysis using 13 waves of annual data. Scand. J. Work Environ. Health 43, 50–58. http://dx.doi.org/10.5271/ sjweh.3608. Nachreiner, F., 2008. Erfassung psychischer Belastung und Rückwirkung auf die Arbeitsgestaltung – Grenzen der Aussagekraft subjektiver Belastungsanalysen [Measuring psychological strain and its effects on work design - The limitations of subjective stress analysis validity]. Angewandte Arbeitswissenschaft 198, 34–55. Niedhammer, I., Malard, L., Chastang, J.F., 2015. Occupational factors and subsequent major depressive and generalized anxiety disorders in the prospective French national SIP study. BMC Public Health 15, 200. http://dx.doi.org/10.1186/s12889-0151559-y. Nixon, A.E., Mazzola, J.J., Bauer, J., Krueger, J.R., Spector, P.E., 2011. Can work make you sick? A meta-analysis of the relationships between job stressors and physical symptoms. Work Stress 25, 1–22. http://dx.doi.org/10.1080/02678373.2011. 569175. Pejtersen, J.H., Kristensen, T.S., Borg, V., Bjorner, J.B., 2010. The second version of the copenhagen psychosocial questionnaire. Scand. J. Public Health 38, 8–24. http://dx. doi.org/10.1177/1403494809349858. Podsakoff, P.M., MacKenzie, S.B., Lee, J.Y., Podsakoff, N.P., 2003. Common method biases in behavioral research: a critical review of the literature and recommended remedies. J. Appl. Psychol. 88, 879–903. http://dx.doi.org/10.1037/0021-9010.88. 5.879. Prümper, J., 2010. SQUAW. Short Questionnaire for Work Analysis, Unpublished paper. University of Applied Sciences Berlin, Berlin. Prümper, J., Hartmannsgruber, K., Frese, M., 1995. KFZA. Kurz-Fragebogen zur Arbeitsanalyse [SQUAW. Short Questionnaire for Work Analysis]. Zeitschrift für Arbeits- und Organisationspsychologie 39, 125–132. Rau, R., Buyken, D., 2015. Der aktuelle Kenntnisstand über Erkrankungsrisiken durch psychische Arbeitsbelastungen [The current state of knowledge about disease risks caused by mental workload]. Zeitschrift für Arbeits- und Organisationspsychologie 59, 113–129. http://dx.doi.org/10.1026/0932-4089/a000186. Rau, R., Henkel, D., 2013. Zusammenhang von Arbeitsbelastungen und psychischen Erkrankungen: Review der Datenlage [Association of work load and mental disorders: review of the data]. Nervenarzt 84, 791–798. http://dx.doi.org/10.1007/s00115013-3743-6. Rau, R., Morling, K., Rösler, U., 2010. Is there a relationship between major depression and both objectively assessed and perceived demands and control? Work Stress 24, 88–106. http://dx.doi.org/10.1080/02678371003661164. Richter, P., Buruck, G., Debitz, U., 2014. Screening Gesundes Arbeiten – SGA. In: Arbeitsmedizin, B.f.A.u. (Ed.), Gefährdungsbeurteilung psychischer Belastung. Erfahrungen und Empfehlungen. Erich Schmidt Verlag, Berlin, pp. 206–210. Richter, P., Debitz, U., Pohlandt, A., 2009. Evaluation of the quality of job design with the action-oriented software tool REBA - recent developments and applications. In: Schlick, C.M. (Ed.), Methods and Tools of Industrial Engineering and Ergonomics. Springer, Berlin Heidelberg, pp. 322–333. Rissen, D., Melin, B., Sandsjo, L., Dohns, I., Lundberg, U., 2002. Psychophysiological stress reactions, trapezius muscle activity, and neck and shoulder pain among female cashiers before and after introduction of job rotation. Work Stress 16, 127–137. http://dx.doi.org/10.1080/02678370210141530. Rugulies, R., 2012. Studying the effect of the psychosocial work environment on risk of ill-health: towards a more comprehensive assessment of working conditions. Scand. J. Work Environ. Health 38, 187–191. http://dx.doi.org/10.5271/sjweh.3296.
Schuller, K., Roesler, U., Rau, R., 2012. Self-reported job characteristics and negative spillover from work to private life as mediators between expert-rated job characteristics and vital exhaustion. Euro. J. Work Organ. Psychol. 23, 177–189. http:// dx.doi.org/10.1080/1359432X.2012.727555. Semmer, N.K., Grebner, S., Elfering, A., 2003. Beyond self-report: Using observational, physiological, and situation-based measures in research on occupational stress. In: Perrewe, P.L., Ganster, D.C. (Eds.), Research in Occupational Stress and Well Being. Emerald Group Publishing Limited, pp. 205–263. Siegrist, J., 1996. Adverse health effects of high-effort/low-reward conditions. J. Occup. Health Psychol. 1, 27–41. http://dx.doi.org/10.1037/1076-8998.1.1.27. Siegrist, J., Starke, D., Chandola, T., Godin, I., Marmot, M., Niedhammer, I., Peter, R., 2004. The measurement of effort-reward imbalance at work: European comparisons. Soc. Sci. Med. 58, 1483–1499. http://dx.doi.org/10.1016/S0277-9536(03)00351-4. Sigrist, J., Dragano, N., 2007. Rente Mit 67 – Probleme Und Herausforderungen Aus Gesundheitswissenschaftlicher Sicht [Pension With 67 - Problems and Challenges From a Health Scientific View]. Hans-Böckler-Stiftung, Düsseldorf. Sjøgaard, G., Lundberg, U., Kadefors, R., 2000. The role of muscle activity and mental load in the development of pain and degenerative processes at the muscle cell level during computer work. Eur. J. Appl. Physiol. 83, 99–105. http://dx.doi.org/10.1007/ s004210000285. Stab, N., Hacker, W., Weigl, M., 2016. Work organization in hospital wards and nurses' emotional exhaustion: a multi-method study of observation-based assessment and nurses' self-reports. Int. J. Nurs. Stud. 61, 52–62. http://dx.doi.org/10.1016/j. ijnurstu.2016.05.012. Tabanelli, M.C., Depolo, M., Cooke, R.M., Sarchielli, G., Bonfiglioli, R., Mattioli, S., Violante, F.S., 2008. Available instruments for measurement of psychosocial factors in the work environment. Int. Arch. Occup. Environ. Health 82, 1–12. http://dx.doi. org/10.1007/s00420-008-0312-6. Taylor, J.B., Goode, A.P., George, S.Z., Cook, C.E., 2014. Incidence and risk factors for first-time incident low back pain: a systematic review and meta-analysis. Spine J. 14, 2299–2319. http://dx.doi.org/10.1016/j.spinee.2014.01.026. Theorell, T., Hammarstrom, A., Aronsson, G., Traskman Bendz, L., Grape, T., Hogstedt, C., Marteinsdottir, I., Skoog, I., Hall, C., 2015. A systematic review including metaanalysis of work environment and depressive symptoms. BMC Public Health 15, 738. http://dx.doi.org/10.1186/s12889-015-1954-4. Theorell, T., Hasselhorn, H.M., 2005. On cross-sectional questionnaire studies of relationships between psychosocial conditions at work and health - are they reliable? Int. Arch. Occup. Environ. Health 78, 517–522. http://dx.doi.org/10.1007/s00420005-0618-6. Tsutsumi, A., Kawakami, N., 2004. A review of empirical studies on the model of effort–reward imbalance at work: reducing occupational stress by implementing a new theory. Soc. Sci. Med. 59, 2335–2359. http://dx.doi.org/10.1016/j.socscimed.2004. 03.030. van den Berg, T.I., Elders, L.A., de Zwart, B.C., Burdorf, A., 2009. The effects of workrelated and individual factors on the Work Ability Index: a systematic review. Occup. Environ. Med. 66, 211–220. http://dx.doi.org/10.1136/oem.2008.039883. Waldenström, K., Harenstam, A., 2008. Does the job demand-control model correspond to externally assessed demands and control for both women and men? Scand. J. Public Health 36, 242–249. http://dx.doi.org/10.1177/1403494807085079. Warren, N., Dussetschleger, J., Punnett, L., Cherniack, M.G., 2015. Musculoskeletal disorder symptoms in correction officers: why do they increase rapidly with job tenure? Hum. Factors 57, 262–275. http://dx.doi.org/10.1177/0018720814547569.
208